查看文章

The study of intelligent agent training is of great interest to the gaming industry due to its wide application in various game genres and its capabilities of simulating a human-like behavior. In this work two machine learning techniques, namely, a reinforcement learning approach and an Artificial Neural Network (ANN), are used in a fighting game in order to allow the agent/fighter to emulate a human player. We propose a special reward function for the reinforcement learning approach that is capable of integrating specific human-like behaviors to the agent. The ANN is trained with several recorded battles of a human player. The proposed methods are compared to other two reinforcement learning methods presented in the literature. Furthermore, we present a detailed discussion of the empirical evaluations performed, regarding the training process and the main characteristics of each method used. The results …